A question about making money in the FOREX market - page 26
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
to:C-4
1. Leave Yusuf alone already. I don't understand the ridicule of him at all. Is he raising his thread? Well, he's trying to get recommendations on how to improve his own turkey. At least almost all his posts are related to the forum topic.
2. Let's not raise the issue of "old timer" and "newbie" or I'll laugh out loud.
3. really want to understand this henna logic. You're spouting some pile of your perceptions where everything is mixed up and mixed up. When you try to make sense of that pile, you chase it down with another pile and it all gets completely confused. It is POSSIBLE that there is some original thought in that pile after all.
So:
Let's not touch "lagging"!
As you wrote above: "The problem of technical indicators and of the technical analysis in general is that they do not change their behavior depending on the object studied....... So why should we trust this indicator when it will give us the same results on obviously senseless data" (C).
Even lower: "The indicators themselves do not determine the future, nor do they find such correlations, but they do classify current changes very clearly. With their help we can identify only those changes, which we want to study, and if they are interesting, construct on them the dependence of the future from the past - this is the trading system" (C).
Contradiction, however! we are not talking about what indicators are for - prediction, prediction, "classification" etc. You above cross out the whole TA or its inducator part, and below you give them the right to live.
Conclusion?
There is no contradiction here. There are a great many indicators. Certain states of them are treated as some unique market characteristics. Let's say that there cannot be overbought and oversold zones on the SB, but if we attach RSI to this chart, then at some moments it will cross the 30% and 70% boundaries telling us about such zones. Does it mean that the indicator is bad or that it should be thrown out to the garbage? Not at all. It just does not correspond to what it shows. We must abstract from conventional wisdom and dogmas and look carefully at what RSI is trying to show. Start with the SB - calculate its frequency of crossing levels there and compare it with the frequency of crossing those levels in real markets. Is the frequency different? Why is it different? Why would it be different? Could it be that the difference between his behaviour on the SB and his behaviour on the real markets is the very thing he is trying to show us? Modify it so that these differences are more clearly visible. At this step you start a conscious process of creating TS, you don't use an optimizer to enter into your model a randomly chosen variable, instead you know what characteristic must be entered and what properties it must have. That's what fa1947 means, blaspheming the optimizer left and right and the whole TA at the same time. Now you thoroughly understand the work of RSI, you have competently "sharpened" RSI to those market characteristics that you need to find.
After that, you are trying to find the relationship between what shows the modified RSI and the events that occur in the future (it is described in detail above). Note, as before, RSI does not predict the future, it does not say anything about the future state of the system, but gives interesting properties for the last period of time. If such correlations are found, you will be a winner.
Without the first step, it is difficult firstly to understand what the selected indicator shows, and secondly, you can not properly configure its work. How will you use it to find what is not on the SB if you don't know what is not on the SB? If the indicator is really bad, then all it does is filter noise and there will be no difference between its readings both in a sterile random walk system and on real market series (see highlighted text in my quote).
Most people make the mistake of thinking they are trying to trade a future fortune. Most people simply approximate the current trend into the future: prices have been rising for the last three days - so they will continue to rise for some period of time. They are trying to replicate the latest market results in the future. They want the same picture that was drawn at the moment to be drawn again, only with them on board. They want a repeat of history. How can a person who wants a repeat of history trade the future?
..... Most people simply approximate the current trend into the future: prices have been rising for the last three days - so they will continue to rise for a period of time. They are trying to replicate the latest market performance in the future. ....
Please clarify what you mean by "economic sense"?
Market terms have an economic sense. Market price, profitability, payback period, profitability, etc.
For example, the market price is the actual price which is determined according to supply and demand for goods, etc.
Most people do just the opposite. They sell on the upside and buy on the downside. This is a medical fact.
There is no majority and there is no minority. Exactly half buy and the other half sell. Some see a rise in price as a further decline, and some see a rise in price. The essence of what is being said does not change.
That's a big one. Shall I put it in the Annals?
Some are good at trade, some are good at the Russian language) No one is all-powerful.
to C-4
theoretically, you have shown a kind of "ideal" and correct finding scheme. But:
1. If I find an effect on a price series and then the same effect on a SB series does it mean that the effect cannot be used? No! It must, can and must!
2. Is it possible that the effect indicated on a price series will also be indicated on a series of SBs and this effect will influence the profit of TC positively in the future? Yes! Absolutely! Why not? The coincidence of this particular piece of the SB series, randomness - whatever.
3. If I detect a stable probability correlation between the set of indirectors and the price behavior, but I cannot express consciously what exactly is the meaning of the resulting indicator of the set of indirectors, should I still use this "magic" number for building my TS? Yes, absolutely!
NS - you load everything possible on the input, and the network, without thinking about the eq meaning, learns from the sample and looks for hidden correlations.
OK, let's say the network has found a correlation. Positive effect in trade, but perfectionism complex keeps you awake - look for ec interpretation of the pattern found. But no more! And not the fact that you will find it! And not the fact that you will find even the slightest logical explanation of the pattern found
Да, я загнал в НС 100 индюкаторов, заранее не зная комбинация каких из них даст эффект! Да, НС нашла комбинацию! И что, смысла торговать ее нет????
You do not know whether this combination is a fit or not. Even if the combination is not a fit, and it does fumble for something, you don't know what it is, and after a while, when the process changes, the combination will stop catching the profit. In the other case, if the rule set is deliberately sharpened to identify the process, when it changes within certain limits it will follow those changes and keep the system from "going stale".
Don't fool yourself. You don't even know what indicators you are using, you don't know what they show, you don't know what exactly you want to find, you are only interested in one thing - profit. To put it simply: you mix everything in one pile and feed it to the machine. Poor NS gives out something that looks like "pie" so it won't get raped again. Reminds everyone of the famous cartoon about how one lazy kid got into a fairy tale. Instead of making pirozhki in Russian oven, was made a hell of a mixture of dough in a bucket and unmade firewood. The reality is that there is only one thing waiting for such systems - a guaranteed failure. The market is not a neural network, it cannot be forced or raped.
You don't know if the combination is a fit or not. ....
That's a big one. Shall I put it in the Annals?
Better my posts. There are already a couple of worthy candidates.
to C-4:
You don't know whether this combination is a fit or not - absolutely right, I don't! But I do know one thing for sure - the system will give you several combinations and if any combination is not a fit, it will also be among them.
Even if that combination isn't a fit, and it does pick up something, you don't know what it is, and after a while, when the process changes, that combination will stop catching profit. - absolutely right. I will tell you more than that - I have never believed, do not believe and will never believe in "eternal" combinations and systems. What I need from the combination is that it would catch profit for some time and a mechanism for calculating the effectiveness of the system. After the efficiency drops, I will move on to the next one.
In the other case, if a set of rules is deliberately sharpened to identify a process, when it changes within certain limits it will follow those changes and keep the system from "going stale". - Possibly. Quite. How is the theory a practical example? Otherwise it's tales of a land of milky rivers and acidic banks.
Don't kid yourself. You don't even know what indicators to use, you don't know what they show, you don't know what exactly you want to find, you are interested only in one thing - profit. To put it simply: you mix everything in one pile and feed it to the machine. Poor NS gives out something that looks like "pie" so it won't get raped again. Reminds everyone of the famous cartoon about how one lazy kid got into a fairy tale. Instead of making pirozhki in Russian oven, was made a hell of a mixture of dough in a bucket and unmade firewood. In reality, there is only one thing waiting for such systems - a guaranteed failure. The market is not a neural network, it cannot be forced or raped. - Yes, I do, among other things. And I am honestly and frankly aware of this and admit it. And I know that at least some of these pies will be quite edible. Absolutely.
Is it possible to build a TC in this way - yes. And it has been proven by decades of practical trading experience. People added numbers, multiplied and divided them without thinking much about their meaning and got more and more indices. Some of these pies turned out to be edible.
An example of a profitable TS built in a "conscious" way?